Evaluating Noninferiority in Clinical Trials with a Hierarchical End Point

AbstractThe win ratio is a method for analyzing a clinical hierarchy of outcomes. It is most commonly applied to randomized clinical trials to evaluate the superiority of an intervention versus the standard of care. Once an intervention is approved, a noninferiority trial may be used to assess whether or not an alternative intervention is as effective. The authors explain how to select an appropriate noninferiority margin in terms of the win ratio, while emphasizing the importance of translating it into familiar metrics (e.g., rate ratios, differences in means) for easier interpretation. This framework is illustrated with a worked example of a hypothetical trial. The authors provide practical guidance on how the win ratio could be effectively used in future noninferiority trials.